M. B. Gavrikov, N. V. Pestryakova, A. V. Uskov, V. V. Farsobina, “Dependence of the quality of recognition of mutual arrangement of average rasters and vectors of bases of learning and recognition”, Keldysh Institute preprints, 2009,019, 29 pp.

This article is cited in 1 paper

Dependence of the quality of recognition of mutual arrangement of average rasters and vectors of bases of learning and recognition

M. B. Gavrikov, N. V. Pestryakova, A. V. Uskov, V. V. Farsobina

Abstract: For the character recognition method based on regression analysis, an approach to the examination of the quality of recognition according to the degree of difference between the bases of learning and received through the modification of recognition is developed. A study of the correlation in the dynamics of average values (rasters and polynomial vectors) as well as the quality of recognition in the process of changing the base of hand-printed digits is carried out.